Re: ATM All Sky Instrument - a comment on reciprocity failure

["Nils Olof Carlin" <nilsolof.carlin@swipnet.se>]

Richard Schwartz wrote:

> Chemical photography suffers from a defect called RECIPROCITY FAILURE. 
> This means that there is a minimum threshold below which photons are not
> recorded.  The consequences of this problem are that cameras using
> chemical photography are build as fast as possible.  This leads to
> complicated schemes for controlling abberations.

Reciprocity failure is mentioned often or always in the context of chemical
photography, but I have yet to see a discussion of the practical effects of
reciprocity failure.

I have very little experience with astrophotography, but I have at least
tried to estimate those effects. With Kodak Ektar 1000 (I found effects of
a similar order with other 400 color films), halving the aperture (one
stop) was compensated by increasing the exposure time by a factor of about
2.3 (measuring the sky background). 
Decreasing the aperture to 1/32 meant increasing the exposure time  a
factor of appr.64. Increasing exposure time from 1 minute to 1 hour meant a
decrease of sensitivity of 50 %, that is one aperture stop. 

I fail to see that *this* loss of sensitivity is great enough to motivate
complicated schemes and costly optical designs. 

The loss of sensitivity caused by going from a fraction of a second (where
nominal sensitivity is given) to one minute I know nothing of, but
apparently this could be successfully compensated for by hypering.

> 
> However, if CCD's do not suffer reciprocity failure, there is no need
> for a fast optical system and the design can be a lot simpler.
> 
> The big problem with CCD's is that they are tiny.  They have a small
> area, so will cover only a small area of the sky, unless the focal
> length is short.  This tends to drive the camera design to a smaller
> size.  What we need is a 5" x 5" CCD.
> 
Or a good 5 cent cigar.

Nils Olof